all photos by the author, except where otherwise noted
In 1980, Roy Sieber's exhibition “African Household and Furniture Objects” (Sieber 1980) first drew attention to the “functional art” objects of Africa, including wooden headrests and stools. Since then, this category of artifact has gradually drawn more interest among ethnic art lovers, collectors, and researchers, although still in modest dimensions.
European explorers and travellers in Africa collected headrests since at least the mid-nineteenth century (Nettleton 2007:100–101). Headrests and stools are, of course, not confined to Africa alone (Dewey et al. 1993), but have seen a particularly rich and varied development there and are found among many ethnic groups (see also Falgeyrettes 1989). In the past decade or so, the new focus on headrests has led to several major exhibitions in African art museums or ethnology museums and to a spate of websites of traders and collectors.
What is a headrest? It is an object, usually wooden, that people rest their head or hairdo on when lying down and carry with them when traveling. It is a mobile item and remains the personal possession of a specific individual. It is used when taking a nap during the day (or, more rarely, at night); to sit on while talking with others, milking a cow, keeping an eye on things while in a restful pose, e.g., during herding; to lie on and gaze at the starry sky at night; and to show in public as a mark of status and group identitty. The item is used by both sexes, but predominantly by males. In many cases in southern Ethiopia and northern Kenya, the males in agro-pastoral societies use them to protect their elaborate hairdo (e.g., the feathered “clay-cap” coiffure) when resting. Occasionally women use them after having “buttered” their hair, but they have less vulnerable hairdos and are more inclined to remain in the restricted space of the home and, as they do not appear in public spaces or travel as much, leave their headrests at home.
Using a basic typology, I here present a survey of head- or neckrests or stools (hereafter called “headrest-stools”) and their presence specifically in southern Ethiopia, showing the variety of forms and their distribution, and reflect on the practical uses and possible meanings of this seemingly simple artifact. Why is it so widespread? What does it stand for? How are aesthetics and practical form combined? I will also briefly address questions of commoditization and of when and why people abandon its use.
The headrest as an “object of value” has remained marginal in art markets. But it has evoked a renewed interest from ethnic art scholars and also from collectors because of growing scarcity and the rediscovery of its aesthetics. There is a persistent concern in comparative art studies with criteria of aesthetics and functions of “daily art” or apparently “utilitarian” objects such as headrest-stools, which often escaped academic attention, compared to the mainstream of the more salient figurative African art objects.
Published studies are scarce (IES 2000, Nettleton 2007, Boyer 2012), and headrest production and use are not yet decisively affected by globalizing cultural influences. I came across these objects over the past two decades during anthropological fieldwork in Ethiopia among people who still use headrests and stools. Their remarkable variety across cultural traditions, both within Ethiopia and beyond, has not yet been adequately explored.
As to interpretation, I continue a recent line found in some previous studies of African “functional objects” (starting, inter alia, with Sieber 1980, Donovan 1988, Ravenhill 1991) and take the headrests out of the domain of “utilitarian appraisal” to reclassify them as aesthetic artifacts, because they combine a number of aspects: functional, but also artistic, social, and rhetorical. Nettleton, in her path-breaking book, included headrest-stools squarely under the category of art because, for her, they have a demonstrable “spiritual content” (2007:9)
In the move to appraise headrests as tribal or ethnic art, the question of authenticity comes up. When an object is classified as “authentic,” it is presumed to be functional within a meaningful sociocultural context of an ethnic group or people, and its artistic value and appreciation are enhanced. But the notion of “authenticity” in the eyes of Western or other ethnic art collectors or scholars is always problematic. The term presumes a fixed image of continuity and genuineness that is denied to certain objects, and not others. “Authenticity” is indeed largely a construct with only a tenuous basis in historical reality (c.f. Nettleton 2007). Nonetheless, I here focus on “genuine” headrest-stools, in the sense that they are known to have been produced and used by members of an ethnic group, regardless of the age of the object, and were not primarily made for sale to outsiders.
INTERPRETING THE HEADREST-STOOL
Bocola, in African Seats (1995), takes a primarily aesthetic view of the object, sidestepping the anthropological critique that a judgment-based aesthetics does not give due attention to the original social and cultural contexts that give objects their significance. Bocola argues that the two aspects are complementary, and that “[t]hrough its aesthetic dimension, material culture secures a community's collective identity, and thereby fulfils an important social function, albeit one of which the community is hardly aware” (1995:12); an aesthetic approach is said to be entirely alien to the societies from which these objects come—it is an aspect of Western society. But a deeper study of the subject would quickly reveal that both contentions are questionable.
People in “traditional” African societies do have sets of criteria—often implicit—that allow them to say whether an object is well-made, appealing, or beautiful. Admittedly, this is not a purely context-free aesthetic appraisal: they often add something that derives from the “life history” of the object such as who possessed it, for how long, and how and why was it acquired. But this poses a new interpretative challenge, interwoven with social relations and cultural values (c.f. Ravenhill 1991:7, Abbink 1999). Bocola's contention that material culture secures a community's “collective identity” should be interpreted with care, because often the same object is produced in several communities and thus refers to the identity of people in more than one group. Objects and styles travel, boundaries between artifact traditions are fluid and will continue to be so in a globalizing world (see Kasfir's seminal study on the erroneous “one tribe, one style” paradigm, 1984:166; also Nettleton 2007:23). In general, the issue of aesthetics and artistry in “tribal”/ethnic groups outside the Western setting has been an issue of long-standing debate among anthropologists and art scholars (Jopling 1971, d'Azevedo 1989, Coote and Shelton 1992, Gell 1998, among the ground-breaking works). Thompson (1989) demonstrated the delicate aesthetic criteria found in Yoruba art, while Fernandez (1989) has emphasized the range of intense artistry in Fang art defined by notions of aesthetic “order.”
The question of artistic agency remains important: in the non-industrial, small-scale societies under discussion, the producers of material culture objects (including those that carry symbolic meaning beyond functionality) are personally involved in and responsible for the final product: either crafting it themselves or finishing it to their own taste. There is obviously no mass production of a standardized commodity. This creates the preconditions of agency, of individual artistry (c.f. Bassi 2000) and expression in this specific object, making it a semiotic item in the production of cultural meaning. This meaning is often hidden from its face value—e.g., for tourists or occasional travelers—by its daily use in a clear utilitarian function.
THE DISTRIBUTION AND DIVERSITY OF HEADRESTS
Seen as a class in itself, African headrests and stools offer an astonishing variety of figurative and nonfigurative types and styles all across the continent (c.f. Bocola 1995:11; Nettleton 2007), although basic forms recur in different regions, independently from each other. Ethiopia is only one region where this object developed, with many variations across the country. The Christian and Islamic areas of highland Ethiopia have their chairs and artful woodworks, but do not know the headrests and stools, and therefore these areas will not be treated here. Today, the greatest variety is found among peoples of southern and southeastern Ethiopia (in the Afar, Boran, and Somali regions and in the Omo Valley area), where the production of headrests continues, although in gradual decline in many areas. Several groups have started using larger wooden stools, chairs, or imported seats/benches instead, slowly abandoning the use of portable headrests altogether.
After some comments on efforts at typology or classification, I will present in more detail a number of southern Ethiopian headrest-stools as prime examples of this genre of functional art. The five “ethnic groups” are discussed on the basis of known models of headrest-stools that reflect the basic principles of the typologies. These groups also have a living tradition of production and use. Most of the headrest-stool types referred to here are new and have not been previously described in catalogues or in the literature.
Typologies of the headrest-stool can be made on the basis of form, gender distribution, function, and social symbolism. Typological exercise is descriptively useful but of limited value due to the diversity of the object. One type I do not discuss: the three-legged stools, or rather, small elongated “benches,” prevalent among several herding peoples in the south, such as Bodi, Nyangatom, or Suri. These stools are made out of a natural formation of at least three tree branches emerging from a trunk and are roughly carved. They are seldom carried around during travel but are domestic furniture, staying near the house. They are also used less by adult men or elders, and more by women, children, and visitors (see Van der Stappen 1996:125–29, IES 2000:29).
Although a unified and fixed typology of headrest-stools cannot perhaps be arrived at, Nettleton (2007) made an important contribution towards a classification of headrest-stool types and styles across Africa. Although not entirely satisfactory, I take her work as a starting point and add to it by referring to Dohrmann's (2010) survey of South Ethiopian headrests.
With some exaggeration, Nettleton noted “a remarkable degree of formal similarity among these headrests, all of which fall within the fairly narrow band of possible structural arrangements” (2007:24). While the function of the object certainly imposes limitations on its form (as to height, size, stability of form, weight, and durability), Nettleton's statement belies the enormous diversity of forms, types, and solutions to the construction challenges of a headrest-stool. Headrests are perhaps only unified in their need to have a base or foot, a support structure (one column or more), and a flat surface or plateau to lie or sit on, along with overall stability of design and more or less durable material.
Nettleton classifies East African headrest-stools on the basis of a) “polygonal support and angular basis” and b) “planar support and domical basis” (2007:204–205). The first type has one or more support pillars under the plateau, with a wider, rectangular base on the ground (Fig. 1). The second type is specifically characterized by a semicircular, reinforced, domelike base and a strong, broad supporting pillar (Fig. 2).1
Dohrmann limits herself to a classification of (Ethiopian) headrests based on the forms of the support leg (column) and the surface, and distinguishes a) the block-wood headrest with concave top (Gurage, some Oromo, Sidama); b) the column headrest, with one or several support columns on a broader foot or pedestal as base (Arsi Oromo, Gurage groups, Sidama, some Somali); c) the headrest with a concave plateau with one conical leg without a pedestal part (Kaficho, some Gurage, some Oromo); d) a headrest-stool with one broader, “winged” plateau—square or elliptic, either convex or concave—with a rectangular support column and a domelike, rectangular, or round pedestal (Hamar-Banna-Bashada, Kara, Suri, Nyangatom, or Mursi; see Figs. 2, 7, 8);2 and e) the three-or-more-legged stools or “benches” made from a natural branch formation of a tree. This type is common across virtually all groups of southern Ethiopia, but these stools are not as highly valued as the headrests proper and are often less carefully produced; they are more often seen as regular furniture.
HEADREST-STOOLS IN SOUTHERN ETHIOPIA
Southern Ethiopia knows a large number of ethnic groups, with varying degrees of distinctiveness in language or dialect and cultural tradition. They also differ in degree of interethnic and regional contacts or wider national integration, ranging from relatively “isolated,” like Suri or Mursi—readily identifiable on the basis of way of life and external appearance—to strongly mixed, dispersed, or integrated, like Kämbata, Hadiyya, Wolaitta, or various Gurage peoples. Often they overlap to such an extent that people's “identity” is difficult to ascertain.
In the social and political domain, the boundaries between various groups were always fluid and we have been long convinced of the need for an “historical process model” (Kasfir 1984:184). Still, their material-culture traditions or artifact styles differ in emphasis and function as a point of reference, both for rural people and for those in urban areas. The headrests are one such a tradition, despite the processes of mutual influencing and the erosion of material culture production fueled by the political and cultural dominance of Ethiopian highland society, persistent insecurity due to intergroup rivalry and fighting, political marginalization and, not least, imports of cheap, mass-produced plastic and iron goods gradually replacing traditional craft work. But these new import objects are also perceived to have advantages: durability, easy procurement, and the evocation of a general sense of modernity (a new connectedness). Cultural disdain on the part of other (highland) Ethiopians towards the people of the south are evident in government policy (Epple and Thubauville 2012) and may contribute to people feeling a certain embarrassment toward their traditional artifacts.
Below, I first discuss headrest examples from Me'en, Suri, Mursi, and Nyangatom: different peoples belonging to the Nilo-Saharan language family. They are mostly agro-pastoralists, with cattle-herding as an important socioeconomic activity, and have “acephalous,” segmentary societies with no central chiefly authority, but often with age- or generation-grades and ritual leaders. In the past century some groups, like Me'en, largely left herding and became predominantly shifting cultivators and farmers in higher-altitude areas.
A second (Omotic-speaking) group comprises Kara, Hamar, Bashada, Aari, Dizi, Maale, Gofa, Malo, and Gamo peoples. These were mixed farmers and herders. Some of them specialized in agro-pastoralism, others developed sedentary agricultural societies with highly complex hierarchical chiefdoms, like Dizi, or with a “divine kingship” structure, like Maale or Gofa. But not all presently have headrest-stools.
Me'en. Me'en are a people (numbering approximately 152,000)3 of mixed agriculturalists and shifting cultivators in the Maji area of southwest Ethiopia. The Tishana branch lives in the highlands west of the Omo River; the agro-pastoral Bodi branch lives east of the river. Bodi4 number about 7,000 people. These two branches of Me'en do not have much contact.
The Me'en headrest-stools (called chákam in their language) are a unique type and mostly have the same style: an oval, concave plateau with a small indentation on the long sides and one sturdy leg, sometimes finely carved into four separate pillars, with a domelike foot and straps variously made of braided cowhide, tree bark rope, aluminium, copper, or metal from old umbrellas.
All Me'en headrest-stools have a double function, but they are predominantly used as stools. The invariable ratio of the two parts is 3:2, i.e., the height is two-thirds of the width of the plateau. The stool's symmetry is deemed very important by the makers and users. Characteristic of these Me'en stools is the plateau, which is slightly tilted forward, probably to make the sitting position more comfortable. Owners take great care to polish this stool very smooth and to give it a deep dark-brown color.
Fig. 3 shows a unique type which used to be carried by Me'en local headmen. It has a wooden handle, cut from one piece of wood. It is a rare object, and very few Me'en admit to having the skill to make this type. I only saw two examples of this type during my fieldwork.5
Fig. 4 is a very commonly used, classical Me'en type, well-polished and very symmetrical, while Fig. 5 is a less well-crafted one marked by heavy use (with a crooked base). It shows the individual touch of the owner in the cow's tail hair decoration near the strap. Fig. 6, finally, shows the more common Bodi-Me'en type called chákam.
Suri (and Mursi). Suri (with two subgroups, Tirmaga and Chai) are an agro-pastoral people living in the southwest of Ethiopia near the Sudan border, counting some 24,000 people. They were until recently an autonomous, politically marginal people in Ethiopia. They live by transhumant cattle-herding and shifting cultivation of food crops. They also do alluvial gold panning. Mursi people, living to the east of the Omo River, are very similar to Suri in language, social organization, and material culture traditions, including headrest-stools. Neither group is marked by high skill in woodwork, and they show significant borrowing from neighboring groups, including in their headrest-stools.
Suri do not seem to have a developed style of headrest-stool that they can exclusively call their own, but one or two types are more common than others, showing affinity with those of neighboring groups (Figs. 7, 8, 10). Fig. 9 (c.f. Fig. 3) is perhaps a characteristic Chai-Suri headrest (nyakachóli). For the rest, they are easy in borrowing the style of their neighbors like Kara and Nyangatom, as seen from the elongated, concave shape of the plateau and the support pillar (Figs. 8, 10), and of the Hamar (Fig. 10). Suri men also take less care in the polishing and fine finishing of headrest-stools (Fig. 9 is a notable exception).6
Suri distinguish the three-legged wooden stool (állé) from the one-column small headrest (nyakachóli). Both are usually made of hardwood species, for instance Grewia mollis or Acacia sp. The állé used to be the more common style. The nyakachóli was probably taken over from the neighboring Nyangatom (c.f. Verswijver and Silvester 2008:241). Among Mursi, on the east side of the Omo River, similar types of headrest are seen, also influenced by Nyangatom and Hamar.
Nyangatom. This agro-pastoral people, counting 26,000 and speaking a Para-Nilotic language, is culturally related to Turkana and Toposa peoples. Nyangatom (or Bume) live on both sides of the South Sudan–Ethiopia border. Over the past two decades they moved gradually northward into the territory of Suri and are on bad terms with them due to multiple armed clashes, cattle raids, and competition for pasture and water sources. A brittle peace has held since 2007 (Sullivan 2008). Nyangatom are a cohesive, age-group-organized society of cattle herders and cultivators, with a material culture quite similar to that of Suri and Dassanetch. Their headrest-stools are called akicholong and come in various types. They are only carried by male adults, and their use is regulated.
A rare headrest-stool from Nyangatom, likely showing Turkana influence, is seen in Fig. 11, made of grewia wood (Grewia mollis Juss.) and has a cow-leather strap. A very common type among both Nyangatom and Dassanetch has a convex saddle seat (Fig. 12; see also Verswijver and Silvester 2008:238). The Nyangatom example in Fig. 13 is akin to the Hamar-type headrest with the saddle shape, but its integrated wooden handle is a feature not seen among the Hamar types. It has stripe incisions at the foot part, again unlike the Hamar type. Nyangatom do not apply the multiple geometric patterns as decoration that are often seen on the Hamar-Banna-Bashada examples, as seen in the plateau of Fig. 14.
Among Nyangatom, the social symbolism of the headrest-stool is very evident: women cannot touch males’ headrest-stools, and it is even said—of course, by men—that when they do, they may become “infertile” (Verswijver and Silvester 2008:239). We see a clear connection with the cultural symbolism of gender relations, masculinity ideals, and reproduction. Also, young male Nyangatom may only start using a headrest-stool after their initiation in the relevant age-set (Verswijver and Silvester 2008:239), as was the case among Bashada or Hamar when their age organization was still intact.
Apart from the tripod stools/benches not discussed here, headrests of southwest Ethiopia predominantly show two types: a) the column headrest, with one or several support columns on a broader foot or pedestal as base, and b) the headrest-stool with one broader, “winged” plateau, square or elliptic, either convex or concave, with a rectangular support-column and a dome-like or rectangular or round pedestal.
Hamar-Banna-Bashada (HBB). Hamar (pop. 47,000), Banna (pop. 27,000) and Bashada (pop. 2,700) live in contiguous territories in the South Omo area and are agro-pastoral subsistence peoples, speaking a Cushitic language. They are culturally and linguistically very similar. One of the main functional uses of headrest-stools among adult males of this group is to keep the head above the ground day and night in order to protect their complex clay-cap head decorations. (Among Me'en, Suri, and Mursi, this custom of head decoration is absent.)
The main type of headrest among HBB is the well-known saddle shape (see Fig. 14), an ingenious form allowing both head repose and sitting. It is also widespread among Turkana. Adult married men in public all carry a headrest (borkotto), which is—more than among other groups—a core part of their social identity. There are rules on their use: women do not handle them; boys cannot use them either. Among Bashada, a young man could only start using one after his age initiation. Men in HBB groups are inseparable from their headrests (Strecker 2000:17), which are an indispensable part of their personal identity and a source of cultural belonging and self-esteem.
The archetypical HBB model is also found among Kara and Nyangatom (Fig. 13), and even among Suri (Fig. 10). Strecker (2000:18–19) claims that its origin lies with Nyangatom and Dassanetch. The decorated Hamar borkotto is popular among neighboring groups, who buy them (SORC 2001:94).
Kara. Kara people (pop. about 2,500) are in an intermediate position, being neighbors of Hamar, Bashada, Nyangatom, and Mursi. They are cultivators and fishermen, with only small numbers of cattle. Perhaps due to their small number and location near more powerful neighbors, they are a prime example of adopting a fusion of styles within one “ethnic group.”
All the types of headrest found among their neighbors are also found among Kara, but they have one specific type of stool (Fig. 15; see also Verswijver and Silvester 2008:240): a sturdy, two-legged form with a central, integrated handle at the lower end. Again, it is both a stool and headrest, only used by adult males (and formerly restricted to initiated males only; Verswijver and Silvester 2008:241). This type was taken over by Dassanetch, agro-pastoralists living north of Lake Turkana (Fig. 16).
Among all southern Ethiopian groups, the production of a headrest is a social activity and a process of patient work by male experts. While many males learn the craft of their production, there are always specialists who are commissioned to make one, because of their better aesthetic insight.
Production takes place in a social setting, often in a group of males talking in the shade of a tree while they discuss personal and community affairs. The carving is not a hasty job and can take a week or more. The first step is selecting a tree of a specific kind, usually with a tough, durable wood. The piece of wood is left above the fireplace to dry. Then the carving starts. When finished, it is left to sit for a week or so. It is then rubbed with leaves and with earth, butter, or animal fat. Among some groups the item is buried in the ground to let it absorb these substances. The next stage is to polish it with sandpaper-like leaves (e.g., of the Lippia grandifolia Hochst. tree), smooth it, treat it with castor oil (of the fruit of Ricinus communis L.), and rub it. Polishing to a deep brown, with additional castor-seed oil, is repeated several times over a longer period. The final step is to apply decoration and attach a leather carrying strap or an aluminium handle.
Aesthetic appreciation differs strongly per headrest and some are rejected as ugly, despite functioning adequately. An example of a beautiful Me'en headrest, according to local standards, is seen in Figure 5, characterized by: a) skillful and delicate carving: the wooden handle carved out of one block of wood, the right symmetry, height and width proportions, and b) the warm, brown color with good patina, due to careful and repeated polishing. Me'en admitted that the skill needed to make this one is becoming rare in today's generation. As noted above, I only saw two examples of this type during my years of fieldwork since 1990.
The square Me'en stool in Figure 17 was made by a novice carver in the 1980s and was not appreciated by his peers: it was considered ugly and not fit to be seen with. The square form of the plateau is unusual (for Me'en) and not pleasing to the eye, the concavity of the top is inadequate, symmetry is weak, the edges are not well-finished, the pedestal at the base is coarse and hastily carved, and the finishing is nonchalant. This one was used only for a short period of time and then given away (to the anthropologist).
THE CULTURAL CONTEXT OF USE: ETHIOPIAN EXAMPLES AND THEIR CHARACTERISTICS
A key characteristic of Ethiopian headrests is that they are not figurative: they do not show faces or human and animal figures; there are no caryatids. All have a geometric or abstract form, in the sense of the shape not being representational or indicative of symbolic meaning in itself. This contrasts with examples from Central Africa, e.g. from Shona, Luba, Yaka, Kuba, and other peoples, where caryatids, recognizable animal figures, or other representational elements are common.7 All the Ethiopian headrests have such abstract or formal shapes, and most even lack decorative drawings. Those with figurative motifs are recent innovations of the last decade or so, with an eye to the tourist market.
Second, the headrest-stool form is always symmetrical. The greater the symmetry, the more it is admired and cherished. This symmetry or balance is also seen in carvers’ efforts to have the pith, the central part of the tree branch or trunk, fall exactly in the middle of the plateau, as can be seen in headrest-stools of Chai-Suri and Dizi (Fig. 18).
Symmetry is an element of harmonious order or balance, which is a subjacent criterion in the appreciation of any art object. Lohse (1974:47) has cited G.D. Birkhoff's formula of aesthetic balance. Birkhoff, a mathematician with a keen interest in aesthetic theory, offered this formula, claimed as a universal one valid across cultures, in his 1933 book Aesthetic Measure: M=O/C, meaning that aesthetic measure or balance (M) equals order (O) divided by complexity (C), i.e., the more ordered, symmetrical, and balanced and the less complex (in terms of unnecessary elaboration and ornamentation) an art object is, the more “beautiful,” apt, and aesthetically rewarding it appears. It applies to the best examples of the headrest-stool as well.8
Third, headrests are predominantly but not exclusively associated with a specific people or ethnic group. For instance, an Arbore man in a discussion on headrest-stools said: “This is the real Arbore stool” (SORC 2001:94), and he added that its use was limited to older men. There is a sociocultural context that explains and enhances their attractiveness or relevance. But many types migrate: they are admired and adopted by people from other groups that like them and reproduce them, or are given by males to their bond-friends in another community. This transethnic experimentation has led to a blurring of boundaries.
It does not diminish their “authenticity,” as headrests were never in themselves designed to mark ethnic or tribal identity. They historically came to be associated with a group due to geographical concentration or proximity. But artistic expertise and achievement—also of specific producers—transcend ethnic group boundaries.
Fourth, the size of the headrests-stools is fairly uniform: rarely above 15–18 cm in height and some 25–30 cm in width, with one or more “legs” and with a platform fit to sit on or to rest the head on. Proportions seem also to be rather constant: a headrest or neck-rest proper has a proportion of roughly 1:1 of the horizontal and vertical axis; i.e., height equals the size or the width of the plateau. Headrest-stools such as those of Me'en or Hamar, most frequently used for sitting, tend to have a 3:2 proportion of parts: the height being two-thirds of the width.9
Fifth, the production of a headrest is a creative, artful act, demanding skill and feeling. Not every male can make an acceptable one. Some are not liked, even rejected. In most societies, specialists emerge because their designs and quality of work are recognized. In sedentary Ethiopian societies like Bale-Oromo, Arsi-Oromo, or various Gurage groups, a lower-rank craftsmen class of woodcarvers (called ogessa, or fuga) made the headrests. Among more egalitarian agro-pastoral peoples, adult men either carve their headrest-stool themselves or give the job to an experienced carver, who has no inferior status but is admired and recommended for his work. The finishing of a headrest-stool is done by the owner.
Sixth, headrests are cherished personal objects, shown in public and transmitting clear indications of the social status, gender, and age of the owner, who is almost always a male adult.10 The personality of the owner is also visible in the final finishing (e.g., decoration, coloring, engraving) of the item, even if he did not carve it himself.
THE PRODUCTION OF MEANING: SYMBOLIC, AESTHETIC OR UTILITARIAN?
The two characteristics just mentioned—the creative act of producing the headrest and its function as repository of male personality—point to the production of meaning in headrest-stools. In the headrest we clearly see a combination of aesthetic and utilitarian functions. The aesthetic aspect is not only in the eye of the outsider, but also in those of the producers and the people (“customers”) themselves. Individual skills and appreciation vary markedly.
The above six characteristics yield clues for answering the question of whether the headrest-stools are a form of art. The debate that followed Donovan's paper of 1988 on Turkana artifacts has led to the conclusion that there is no doubt artistry is involved in African “functional art,” due to individual artists’ creativity, aims for beauty and aptness of form, and references to a transcendental dimension of valuation. We could say that this transcendental dimension is absent in ethnic functional art, or at least is not comparable to that in Western art since the Greeks or to Ethiopian-Christian religious paintings—reflections of static ideals of perfect form and essence, cosmic or divine order, of human perfection, and of destiny or ideals of redemption and heaven. But in the headrest-stools and other ethnic/tribal artifacts, there is tacit meaning—and not only the surface meaning of a “well-crafted” object.
Any meaning or reference beyond the utilitarian use of a headrest makes it a cultural-aesthetic object as well as a semiotic referent (for something beyond lying/sitting on it). Thus, as Sieber said, “the utilitarian object can, through the successive levels of meaning that may be attached to it, arrive at a far remove from ‘ordinary’ furniture” (1980:125).
A lot has indeed been said of the “symbolic” meanings of the headrest, but often commentary tends toward psychological over-interpretation, based on insufficient knowledge. The object may have deep symbolic referents in some groups, but this is by no means universally the case. A typical comment on a website selling the headrests is the following:
These neck-stools have magical functions, or to be more exact, are medium-like. […] In Africa, this object is generally considered as an intermediary force between the world of the dead and that of the living. Because of this, it is a strictly personal item, as a transmitter of dreams and thoughts that must then be interpreted [by the owner].11
This is more like a sales pitch, creating an aura around the object and thus framing value. The dream-transmission function is speculative, and while it may be true for Melanesia (Nettleton 2007:9; c.f. Dewey 1993:180, Falgayrettes 1989:108–109), for Africa the “symbolism” and the “magic” or spiritual functions of headrests are not widespread, although they might be evident in a few cases, like Ashanti stools, kisumbi stools among Lega in the Congo region, Shona in Southern Africa (Nettleton 1990) or among Luba, where figures in caryatid stools can symbolize the spirit world (Roberts and Roberts 1996:23, 29). It is not so in southern Ethiopia (Boyer 2012:86). The headrest is here seen primarily as a means of repose, of restful alertness, and as an adult male's required status attribute and social marker; not as a connection to dreams or ancestors.
On the basis of the six characteristics noted in the previous section, I claim that the utilitarian, symbolic, and aesthetic aspects of the headrest-stools among most of the Southwest groups under discussion are integrated and convey the object's role as an “extension of the self,” in that they fuse smooth functionality as seat and headrest with beauty and aptness of material form into an aesthetics of personal (not primarily bodily) identity that the (male) owners aim for and want recognized in public. In this respect, Jacques Derrida's well-known argument on prosthetics (1996) may not be applicable here: prostheses may start as physical extensions aiming to rectify or negate “physical handicaps” (the basic meaning is “replacing a lack,” says Derrida [1996:921]), and that may indeed become psychological extension. But I emphasize the persona aspect, the identity issue—present ab initio—in the use of these personal headrest objects, which publicly indicate or “elevate” the user/owner's status. The headrest-stools express a material rhetoric, asserting social position and “competence” in an aesthetically rewarding manner.
THE CONTEXT OF CHANGE: HEADREST-STOOLS IN A STRUGGLE FOR SURVIVAL
The headrest-stool is also an object in transition: there are processes of decline as well as transformation and renewal going on. Among many groups in southern Ethiopia—and also among Somali, Gurage, or Afar—they are maintained and produced for local use, not for tourist markets and are still connected to a cultural context and way of life. They have implicational meanings for gender relations, personal identity, and cultural self-esteem. At the same time, many groups have already given up on the tradition, voluntarily or not. As the integration of local societies into wider Ethiopian society proceeds, men change their focus of identification because of education, (out-)migration, and changing economic activities, which bring them into contact with other groups and urban settings. Also, as societies become more sedentary and less transhumant, mobility patterns change and men no longer travel alongside herds with their personal headrest-stools. People settling in more fixed hamlets or villages adopt larger stools or chairs that stay in or near the house. In addition, religious change may play a role. Conversion to (Evangelical) Christianity, for instance, often leads to a rejection of indigenous culture, including material culture.
Finally, there is the impact of the cultural prestige economy in Ethiopia: policies and discourses of cultural disdain emanating from highland society, and state presence, lead to a downgrading of indigenous material and spiritual culture. This process is enhanced by the import of the relatively cheap, new objects that are bought for their functionality and durability (plastic and metal objects, containers, and furniture like benches, chairs, etc.). Growing intergroup contacts and imposed socioeconomic schemes (e.g., via strong anti-pastoralist state policies) do the rest. A few nice specimens of lost material culture traditions are the following, from three ethnic groups in the South.
Dizi. Dizi are the sedentary agricultural neighbors of Suri and Me'en, ancient settlers living in the cool Maji highlands, counting some 37,000 people. They belong to the Omotic language group and have bad relations with the neighboring Suri, caught in a pattern of raiding and clashes that has cost the lives of hundreds of people in the past decades. Dizi traditionally had a strongly hierarchical society, ranked in six or seven status groups, with chiefly groups at the top.
Figure 19 shows a rare object from a Dizi chiefly family, at least forty to sixty years old, made in the time of Emperor Haile Sellassie (1940s–50s). It was used by a chief of the higher rank (kyaz) and could not be touched by people of low rank, like peasant serfs or leather craftsmen. The stool is made from tough lowland wood and has a cowhide strap. It seems inspired by the Suri model (see the small base protrusions) as well as by the Me'en (the shape of the plateau). Dizi chiefs in the past had ritual relations with leading Suri families, and objects like headrests may originally have been exchange gifts.
Dizi now have stopped producing this object, the history and meaning of which is not documented. They adopted the highland-type three-legged “Jimma”-type stool, which is much bigger and is seen all over southern Ethiopia in various forms.
Maale. Maale are a South Omotic language-speaking group of mixed agriculturalists and cattle herders, living north of Hamar and Aari people east of the Omo (about 99,000 people) in small villages spread out across the countryside. They had a ritual kingship (kati) and have a local reputation of being skilled, productive farmers and self-conscious, independent people.
The unusual model in Figure 20, never seen elsewhere, has a convex “saddle” structure turned up at the sides and is very well-proportioned, with perfect symmetry and a good, dark-red patina. The carrying strap is of hippopotamus hide. This specific model is taller than the average headrest (22 cm) and has a width of 20 cm, the height perhaps to accentuate the importance of its owner: it belonged to the last kaati of Maale to function in Balla village. This bokoto-type headrest is still made and used by Maale men, but this specific form is no longer produced. Instead, older Maale men use an oyta, a stool with a circular plateau and three or four legs.
Silt'e. Finally, we see a special type from Silt'e people (numbering around 950,000), living around the town of Worabé and throughout urban Ethiopia (Figs. 21–22). The Silt'e, a people of traders and agriculturalists, were formerly reckoned as part of the Gurage people but now have claimed independent status, recognized in a separate political Zone. With the growing Islamization of the Silt'é in the past thirty years, the headrest has been completely given up. It can occasionally be found in Addis Ababa souvenir shops.
COMMODITIZATION: ENTRY OF HEADRESTS INTO GLOBAL TOURIST MARKET FORCES
In the wake of processes of change, head-rest-stools are also being commoditized. The object is detached from its context of use and made into a sellable item, entering different spheres of valuation, notably an economy of consumption. As Steiner has noted in his pioneer work on the commoditization of African art (1994), ethnic art objects are transformed in meaning, function, and even appearance when removed from their context of origin via channels of sale and trade. The aesthetics and meaning of headrest-stools are redefined with reference to the global market arena.
For several decades, Addis Ababa souvenir-shop owners have been offering Ethiopian headrests, stools, and chairs—among other items—for sale. They have supply networks of agents across the country who spot objects and bring them to the capital. Traders actively play into tourist and connoisseur demands for something new, which has led to an upsurge of sales of headrest-stools since the late 1980s. Hundreds of items are on display in the crammed shops, showing a mixture of genuine examples, used and locally bought, and many new ones, made in studios on the basis of old models. The dealers also have connections to global traders and customers, and hence several hundred items now appear in online shops and in collections. Many leading ethnological museums the world over have built up good collections of headrests and stools over the past decades and display them with an eye to their aesthetic value.
In view of the growing scarcity and the tourist-consumer desire for variety, new headrest models are designed and marketed. Despite the academic critiques of the concept of ethnic or tribal “authenticity,” there is always a concern with questions of origins and use, and some new models can be said to lack a basis in “traditional” culture: many are simply not known or used in the societies they allegedly come from. They are produced by urban “artisans” in the backstage ateliers of urban souvenir shops. It is hard to call them “authentic” in their contexts of use.
As such, headrest production is certainly a living tradition, but with the usual measure of invention arising from the need to cater to the demands of the tourist market. Local entrepreneurs and souvenir-shop owners give woodcutters ideas (and photographs or drawings) to develop new forms, which are then often presented as new “discoveries,” and the familiar techniques of giving them a user patina are employed. In this section I present a few examples.
Figures 23–25 show types that are not authentic in the sense that they neither represent traditional forms found among the ethnic groups where they allegedly originate, nor were ever in actual “tribal use.” While the forms are attractive and even original, they were unknown to informants from these ethnic groups. But to many outsiders they indeed “look traditional,” and tourists inquire no further and buy them.12
Figure 23 is an example of an invented, creatively modified headrest: a “Gurage” example (although often labeled as “Oromo” as well) on the basis of an old model, but with a new kind of religious picture and without genuine patina. The original headrests in use carried no such pictures. We see here a hybrid product combining a traditional, nonreligious object with Christian religious symbolism, which plays into the tourist image of Ethiopia as a predominantly Christian highland country. In some online shops, remarkable prices are asked for such items.13
Another one is the so-called Kaficho or Arsi style, with metal or silver decorations (Fig. 24), without traces of use. Kaficho elite headrests often had colorful bead decorations (as can be seen on Fig. 25, collected by French traveller Jules Borelli in the 1880s). But this kind of decorative metal embellishment was not found until some fifteen years ago.
A final one is an alleged “Mursi” headrest, a completely new type as to form and coloring (Fig. 26). This type is never seen in Mursi country, but only in Addis Ababa, on sale in souvenir shops and other tourist outlets since 2010. Photographed in a shop in Addis Ababa in 2010, it has no signs of use either.
Commercialization of headrests is thus well in progress, and follows a trajectory passed by many other “ethnic art” objects before (Steiner 1994). The process further blurs the distinctions between utilitarian/functional art used by members of societies, and specially produced, marketable ethnic art (or kitsch).
Due to their migration into the domain of global markets, the “meaning” and valuation of headrests is of course altered. Their inclusion in a sphere of extraversion outside the societies of origin creates new frames of reference relevant for the acquirers of the objects, i.e., within a status economy elsewhere. Obviously, the objects no longer reflect the sociocultural reference points of social position, gender, status (male adult identity and self-esteem, etc.) or possible links to ancestors or to the cultural past as expressed within the moral economy of the groups of origin. But nevertheless, the objects do convey shades of this cultural meaning when they appear in the display economy of tourists, who proudly exhibit and comment on these objects when returned home.
HEADREST-STOOLS AS “EXTENSIONS OF THE SELF”
Headrest-stools are not autonomous art objects produced and admired for their self-contained beauty alone. This may occur with particularly well-made and patina-rich examples, but the specific conception of this artifact in southern Ethiopia is one of an integrated utilitarian, aesthetic, and symbolic (status-marking) nature. A well-made and well-kept headrest reflects, or extends, the owner's worth and adult status, demanding respect and being taken seriously. It reflects self-confidence in the cultural and material sense, underlining the extension of the owner's self—and not only in the literal sense of having something to sit or rest on. Boyer (2012) has called headrest-stools an “ostentatious” art, but this would be something of an exaggeration: they are assertive of personal identity but there is no competitive display involved and no showing off; they rather exude the quiet self-confidence of the carrier. Here is one of the differences with the culture of masks: no hiding or transforming the self, but expressing or showing it openly.
Regarding the “functional art” items of East Africa and Ethiopia, we can consider the question of “art” versus “handicraft” as no longer in need of debate: there is a definite artistic component in headrest-stools, as also stated by the local makers and users whom I met, and artistic valuation goes beyond the mono-cultural variety rooted in the Western art tradition. In the case of southern Ethiopian ethno-arts, specifically the headrests-stools, a simple but effective fusion of daily functionality and sociocultural referents and standards and an implicational meaning of the object are produced. We see individual artists’ creativity, aiming for beauty and aptness of form, and references to a transcendental dimension of valuation beyond mere functionality. One could say that the good headrest-stool has rhetorical power within its societal setting: via a happy combination of apt form and function, the object aims to draw aesthetic attention and to convince others, so to speak, of the owner's taste, social position, and esteem as a socially adult male person of standing and personal independence. In this sense, the headrest-stool—especially an aesthetically rewarding specimen—is the “extension of the self” of the owner, who thereby is confirmed as a competent and assertive social persona.
A first draft of this paper was presented at the 9th International Conference on the History of Ethiopian Art and Architecture, September 2–6, 2013, Vienna. I thank the participants for their critical remarks and questions. The paper was rewritten during a 2014–15 Fellowship at the Netherlands Institute of Advanced Studies (NIAS), Wassenaar, the Netherlands, which is gratefully acknowledged. All photograph unless otherwise indicated are by the author from objects gathered in the field, or from a private collection of a friend and souvenir trader in Addis Ababa who wished to remain unmentioned.
Nettleton's classification does not cover all Ethiopian headrests: e.g., Sidama or Gurage types are more like wood blocks, or have a carved “wall” as support column.
All population data are derived from the 2007 Ethiopian census.
They are predominantly cattle-herding Me'en (Mela and Chirim sections).
Fieldwork in the Me'en area was carried out in 1989–90, 1995, 2000, and December 2011.
Suri males seem to have shifted their care for a personal male object like the headrest (nyakachólé) to weapons: since the early 1990s semi-automatic rifles are probably their most cherished personal possession.
As demonstrated in neuropsychological research (Sasaki et al. 2005:3159), the human mind is predisposed to appreciate symmetry. This human universal is at work in the appreciation of architecture, (portraits of) the human face, and also headrest-stools.
Somali headrests, on the other hand, which have a cattle-horn-shaped plateau, mostly have a proportion of 2:1.
Among Boran, Arsi and Gurage groups, women do have headrests, but these are used in the private domain, at home, to protect a new hairdo with butter and braiding from dust and dirt, and to retain its shape.
“Ces appui-nuques remplissent des fonctions magiques, ou plus exactement médiumniques. […] En Afrique, cet objet est généralement considéré comme un intercesseur entre le monde des morts et celui des vivants. Il est de ce fait strictement personnel comme passeur de rěves et de songes qu'il appartient ensuite d'interpréter.” See http://www.artethnique.com/catalogue.php?id=22, accessed November 30, 2013.
For more new, fanciful headrests on sale, see www.artethiopien.com/en/11-ethiopian-headrest?p=2 (accessed July 2, 2013). Notably, the “Guragué” ones on display were never used by the Gurage people.
Courtesy of www.artethiopien.com/fr/appui-nuque-ethiopien/61-appui-nuque-oromo-wellega.html (accessed July 7, 2013). The item in Fig. 24 had a wrong ethnic attribution on this site.